projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: bda0765) | patch
Disable the "call to immediate" optimization on x86-64. It is
authorChris Lattner <sabre@nondot.org>
Wed, 18 Mar 2009 00:43:52 +0000 (00:43 +0000)
committerChris Lattner <sabre@nondot.org>
Wed, 18 Mar 2009 00:43:52 +0000 (00:43 +0000)
commitff81ebf7586fa7992353ed44a8c9d564130fb734
tree27221c42cf2a4d0b823fdb9dec27b2ef3a4acbcetree | snapshot
parentbda0765e0763cd7d9b0980328fc1d718a6773628commit | diff
Disable the "call to immediate" optimization on x86-64.  It is
not safe in general because the immediate could be an arbitrary
value that does not fit in a 32-bit pcrel displacement.
Conservatively fall back to loading the value into a register
and calling through it.

We still do the optzn on X86-32.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@67142 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/X86/X86Instr64bit.td diff | blob | history
test/CodeGen/X86/call-imm.ll diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors